INSTRUMENTS FOR THE MEASUREMENT OF COLOR

Transcription

1 INSTRUMENTS FOR THE MEASUREMENT OF COLOR Introduction By Dan Randall Datacolor International Charlotte, NC Color is defined as the sensation experienced or caused by light reflected from or transmitted through objects. In the strict sense, we cannot directly measure perceived color, however we can measure and subsequently calculate certain factors which are responsible for producing this sensation of color. The quantification of the color properties of textile materials is of great economic value in industry and instruments are employed to some degree in almost every textile operation involved in textile coloration. Color instrumentation has experienced a tremendous advancement in technology during the past 40 years. In the previous publication of Color Technology in the Textile Industry, Roland Derby 1 stated that in 1954 there were no more than 10 instruments in common use in the textile industry. Today there are several thousand in use throughout North America alone as these instruments have become indispensable in areas such as quality control of processes, quality assurance of finished products, color formulation, and color sorting of piece shipments. During this 40 year growth in color measurement, the instruments have become more accurate, reliable, flexible, smaller, and faster than their predecessors, at significantly lower cost to the user. The variety of designs and features available to the prospective buyer can be overwhelming. This paper is written and dedicated to providing a series of practical guidelines for the technician, colorist, or manager, to better understand the basics of color instrumentation with the hopes that the instrument chosen will meet the requirements demanded by the task. Historical Perspective The first devices for measuring color were absorptiometers which were used to determine by visual inspection whether two solutions were of equal color. This is very similar to holding two glass cylinders of dye solution up to a light and judging whether they are of equal strength and shade, except that the absorptiometer provided a method of adjusting the thickness or path width so that this change in width could be read from a scale. In measuring reflected light from opaque materials such as textiles, the first instruments were reflectometers developed around These early instruments were designed to closely simulate the visual process as depicted in Fig. 1 lamp Object eye Fig. 1 Visual Process requires light source, object, and a receptor (eye) Three colored filters, Red, Green, and Blue were used to directly measure three Reflectance factors which ideally fit the CIE Standard Observer functions for a given Standard Illuminant, usually Daylight C or D65. Soon afterwards, the reflectometer was further refined to provide output of the tristimulus values X,Y,Z at which point it became known as a Tristimulus Colorimeter. One of the most widely used colorimeters in textiles is the Hunterlab D-25 Color Difference Meter. The beginning of reflectance spectrophotometry dates back to 1928 when A.C. Hardy 2, Professor of Optics at M.I.T. began a project to produce the first spectrophotometer specifically for the measurement of 1

2 reflectance. Commercialization of Hardy s design began in 1935 when General Electric introduced the General Electric Recording Spectrophotometer (GERS). This instrument became the much needed reference spectrophotometer and provided the basis for modern industrial color measurement. The elements of the original Hardy are essentially the same as those used today even though dramatic changes have taken place in their design. As in the visual model, the basic elements of the spectrophotometer are the light source, the object being measured, a means of dispersing the light, and a detection system as shown in Fig. 2. lamp prism detector slit Applications for Colorimeters Since colorimeters are overall simpler to build than spectrophotometers, they are usually lower in cost. As such, they are commonly used as quality control instruments in applications such as color difference, strength determination, fastness determination, shade sorting, to name a few. The cost advantage and simplicity must however be weighed against several serious disadvantages. Firstly, the colorimeter measures the tri-stimulus values for one illuminant and one observer. As such, it is not possible to detect and quantify metamerism. Thus in practice, the colorimeter is used in areas where the standard and the measured batch are non-metameric such as in checking production batches against a production standard made with the same dyes. lamp X Y Z photo-detectors object Fig. 2 Key components of a spectrophotometer - lamp, object, dispersing mechanism, and detector a r g b 45 sample filters Colorimeters. As the first measuring devices, colorimeters were crucial in the development of the science of color, or colorimetry. A colorimeter is a device of fairly simple design based upon the visual concepts of color. The sample is illuminated at a 45 angle relative to the perpendicular line to the plane of the mounted sample. The reflected light is measured directly perpendicular to the sample through a series of three and sometimes four colored filters which represent the relative amounts of red, green, and blue light reflected from the sample. More specifically these filters are designed to ideally simulate the three functions, x,y,z for the Standard Observer so that the instrument directly measures the three Tristimulus values X,Y,Z for the specific illuminant being used. This design is shown in Fig 3. Fig. 3 Diagram of tristimulus colorimeter Spectrocolorimeters A spectrocolorimeter is somewhat of a hybrid instrument which is capable of providing colorimetric data such as X,Y,Z or CIEL*a*b* values for various standard illuminants. In this regard, they are more capable qualtiy control instruments than colorimeters. They are generally priced only slightly higher than tri-stimulus colorimeters, but less than most spectrophotometers. This price differential has been practically eliminated in today s spectrophotometers, and as a result the spectrocolorimeter does not enjoy the niche between the two. The spectrocolorimeter is by design, a spectrophotometer except that it does not output spectral data (%R) at the 2

3 various wavelengths. These instruments are almost exclusively used for applications of quality control. Spectrophotometers Spectrophotometers differ from colorimeters in that they measure reflectance, transmittance, or absorbance for various wavelengths in the spectrum. In the case of reflectance measurement, the quantity measured is termed Reflectance Factor and is defined as the reflectance of the sample at a given wavelength compared to the reflectance of the perfect diffuse white measured under the exact same conditions. This is expressed in the following equation: sample lamp sphere mirror reference beam specular port transmission cell sample beam ref. spectrometer computer sample spectrometer RF(λ) = R(λ) (sample) / R(λ)(pwd) Commonly expressed as a percentage, %R, the reflectance factors are usually referred to as simply % Reflectance. In the measurement of transparent materials such as dye solutions and films, the quantity measured is Transmittance, usually expressed as %T. This quantity is equal to the percentage of light, at a given wavelength, transmitted through a given thickness, usually 10mm, of the sample compared to the light transmitted through the same path without the absorbing sample in place. This may be written as: %T(λ) = T(λ) (sample) / T(λ)(reference) x 100 In practice, the %T(ref) is measured by standardizing the instrument with only the solvent in the glass cell or cuvette. When using reflectance instruments for transmittance measurements, it is essential that the reflectance port be covered with the white standard. Designs of Spectrophotometers All spectrophotometers must have certain key components - Light source, method of spectral separation or dispersion, and a detection system. As a fourth component, most all instruments have a microprocessor on board for data handling and computations. The positioning of these elements and their mode of operation determines the optical geometry of the instrument as shown in Fig. 4. Many of the earlier reflectance spectrophotometers such as the Hardy were designed in a similar fashion to UV/VIS absorbance spectrophotometers used for chemical analysis of liquids in that they employed a scanning mechanism. This provided wavelength by wavelength measurement and data collection at each 1nm or lower if desired. Although extremely accurate, these instruments were slow, mechanical, and expensive. Fig. 4 - Block Diagram of dual-beam spectrophotometer Since reflectance curves are relatively smooth, it is generally agreed that for most applications it is not necessary to measure at 1 nm increments. For this reason, most modern reflectance instruments measure a band of a certain bandwidth which may be 5-20nm in width. Instruments of this type are referred to as abridged spectrophotometers. Instrument Geometry The C.I.E. 3 specified four geometric arrangements for instruments used to measure color. These are (a) 0/45 (b) 45/0 (c) 0/Diffuse and (d) Diffuse/0 as shown in Fig (a) 0 / 45 (b) 45 / 0 (c) 0 / diffuse (d) diffuse / 0 Fig. 5 Recommended C.I.E. Instrument Geometries 3

4 The first angle given is the angle of illumination relative to a perpendicular drawn to the plane of the sample to be measured. This perpendicular is the normal angle, or 0 deg angle. The second angle is the viewing angle again expressed relative to the normal angle for the sample being meaured. The term diffuse is used to indicate that the illumination or viewing is not directional but is rather diffuse, usually by the use of an integrating sphere. While these are the official C.I.E. recommended geometries, a great deal of variation is allowed in commercial instruments. When A.C. Hardy built his first spectrophotometer he found that the surface texture of textile samples lead to poor reproducibility in measurement. As a result, he developed an integrating sphere which provided diffuse illumination thereby reducing the variability due to surface texture. In many 45/0 instruments today, this problem has been resolved by the use of a circumferential ring used either in the illumination or in the detection mode as depicted in Fig 6. This geometry is termed 45/0 Circumferential to differentiate it from the bi-directional 45/0 geometry. spectrometer illumination ring determination of color change such as fastness and staining testing. It is often said that a 45/0 instrument measures not only color difference but also some attributes of appearance such as surface gloss because of it s directional illumination(45/0) or viewing (0/45). While the instrument does not directly measure these geometric attributes, it is no doubt more sensitive to surface texture as is illustrated in the following example: Take two samples A and B which are printed on the same small flat-bed machine. To avoid discussions of pigment printing density and penetration, we will print both samples with acid dyes using the same dye mix however, sample A is printed on a very low gloss (delustered) nylon, whereas sample B is printed on a highly glossy nylon. Now a 45/0 instrument will measure a fairly large color difference (2-3 de CIELAB) because the glossy substrate will give much higher reflectances but lower chroma or saturation. On a diffuse/0 instrument with the specular component included, the same samples will show little color difference (< 0.40 de CIELAB) because the diffuse illumination creates such multiple reflectances that the effects of the gloss are minimized. The question then becomes What do you really want to measure?. For this reason, most instruments for color formulation are diffuse/0 since the colorist wants to measure strictly color, especially when standards are often not dyed or printed on the same substrate as requested for the match. Likewise, in many inspection areas, it is necessary to verify both the geometric quality and color, and in these cases a 45/0 or 0/45 will provide the best assessment sample Fig. 6 45/0 Circumferential Geometry 45/0 or 0/45 Instruments Instruments utilizing 45/0 or the 0/45 were the first to be developed and are believed to closely represent the visual viewing conditions, especially in a light cabinet. There is considerable debate that in most room, office, or retail environments, the illumination is rarely directional, but is rather more diffuse. Instruments with such directional geometry are most widely used in applications of quality control such as pass/fail determination, color difference, shade sorting, or Instruments with Diffuse Geometry Practically all Diffuse / 0 (or 0/ Diffuse) instruments are not truly 0 degree instruments, but are closer to 6-8 degrees off from the normal. This is done to allow for the inclusion of a specular opening within the integrating sphere. The specular component of reflectance may be excluded, although not entirely, by allowing a portion of the gloss to escape through the specular port. The efficiency of this specular port is determined by the overall gloss of the sample and the size of the port relative to the size of the sphere. Measurements with sphere instruments are then designated as either Specular Included (SCI, SPIN) or Specular Excluded (SCE, SPEX). In textile formulation, the normal mode is to measure with the specular included, however in cases where the standard is a glossy paint chip, then better results are obtained by excluding the specular gloss. 4

5 The integrating sphere may be, in theory, of any diameter provided the sample port is not more than 10% of the total area of the sphere. Bench-top instruments usually have a 3-6 diameter, whereas a portable may use a sphere as small as two inches. Their sole purpose is to create illumination which is uniformly diffuse at the point at which at sample is placed. The inside of the sphere is coated with multiple layers of Barium Sulfate which is highly reflective (>90% Reflectance) at all wavelengths. Despite it s high reflectance, Barium Sulfate is not ideal in that the coating is not extremely durable and tends to yellow over time. Instruments with double-beams can compensate for this loss in sphere efficiency, however re-coating the sphere is advisable every few years. Light Sources in Instruments For non-fluorescent materials, the reflectance factors are independent of the illumination (lamp) since they are ratios to the reflectance of the perfect white diffuser (PWD) under the exact same illumination. The only requirement is that the lamp possess sufficient radiant energy throughout the visible spectrum. There are in general two types of lamps used in instruments - tungsten filament and xenon discharge lamps. The early instruments used tungsten filament, usually filtered to simulate daylight. Modern filament lamps are quartz enveloped with a halogen to provide a very stable and intense illumination from nm. This continuous stable illumination was used extensively in single beam instruments such as the Hunter D53, and the ACS Spectro-Sensor. The lamps are very inexpensive but do not last more than six months under normal conditions. There are however, some disadvantages with the tungsten lamps which have contributed to the recent increase in xenon lamps. Continuous tungsten lamps create heat and must be cooled. Secondly, the heat and continuous light exposes the sample which may lead to variation in sample measurement due to such sensitivity. The lamps are usually equipped with infra-red absorbing filters and some models provide a shutter to open only when measuring. In practice, the user should minimize the amount of time a sample is exposed at the measurement port. For these reasons, some manufacturers use pulsed tungsten illumination, however these instruments must also be designed with dual beam optics to account for illuminant fluctuations. The spectral distribution of tungsten filament, xenon, and Daylight are shown in Fig 7. Although tungsten has adequate energy in the visible spectrum for most color measurement, it has much lower energy in the UV-Violet region compared to daylight. This could be a disadvantage when measuring white samples treated with fluorescent whitening agents where sufficient UV energy is needed to excite the fluorescing agent. For this reason, some instruments have been fitted with a secondary UV source such as deuterium to achieve neardaylight illumination. Xenon Discharge Lamps Xenon lamps have been in use since the 1970 s in instruments made by Kollmorgen (Macbeth), Zeiss, and Datacolor. Xenon has many advantages and a few disadvantages. Among the advantages, xenon is a good daylight simulator as shown in Fig 7. In the UV region, un-filtered xenon is much higher than daylight (D65) and usually requires the use of a UV filter to approximate daylight. If left un-filtered, xenon may over-excite a fluorescent material, therefore most all instruments today use a low wattage xenon lamp, or provide a means of filtering the UV portion ( nm). relative energy D65 filtered xenon tungsten wavelength nm Fig. 7 Relative Energy Distributions for D65, xenon, and tungsten filament lamps. Xenon is an inert gas which when highly charged will convert the electron build-up to photons, emitting a flash for a fraction of a second. A sample being measured is therefore not exposed to continuous light, nor is there any heat to dissipate. Although the lamp is intense, it is not as spectrally consistent or stable as a continuous lamp such as tungsten. For this reason, all instruments which use xenon must be dual-beam designs. A reference beam, usually aimed at a point inside the integrating sphere, provides a reference measurement 5

6 against which the sample measurement is adjusted to account for any change in the illumination. Light Dispersion - Filters and Gratings The earliest records of experiments involving the separation of light into spectral colors were those of Newton 4 in 1730 when he used a prism to separate sunlight into the seven spectral colors or bands. In today s instruments, there are primarily two types of dispersing elements used - gratings and filters, with gratings being the most commonly used. It must first be pointed out that the quality or performance of most dispersing elements such as filters and gratings is determined by its ability to separate light into bands of colors. These bands or spectral distribution are measured in nanometers across the width of the individual band (depending upon the detection type) at the point of detection. The width is determined at 50% of maximum peak height for the band measured. Interference Filters The interference filter is very common in instruments, especially those produced during the early rise of industrial color matching in the 1970 s and 80 s. The interference filter is mounted as a filter wheel which is usually rotated by a small electric motor directly in line with the sample and/or reference beam. This simple design uses a single photodiode detector which measures the dispersed light as the filter rotates resulting in bands of variable width. Most are designed to provide an average rather than a fixed bandwidth of about 10nm. Many instruments still in use today are based upon this interference filter such as the Hunter D53, D54, and the ACS Spectro-Sensor, and the ACS Chroma-Sensor 5. Interference filters may also be positioned statically in sequence to provide the necessary spectral distribution. The resulting bands are measured with diode array detectors situated accordingly and are usually nm in bandwidth. Instruments of this type are the X-rite portable spectrophotometers such as 968, and SP series. Diffraction Gratings The first grating was produced in 1821 by Joseph von Fraunhofer and is the most common light dispersing mechanism used today in high performance instruments. A grating is essentially a glass plane with a large number of grooves etched or ruled into the surface. When light strikes this grating, a pattern of diffraction and interference will cause light of different wavelengths to be produced at various angles. A ruled grating with about 300 lines per millimeter will produce a distribution of visible light suitable for measurement. The earliest gratings were of the plane type in that they were made using flat glass and etched. These gratings produced a distribution which was detected by placing photo-diodes along the distribution at certain bandwidths, usually 10nm or 20nm. The plane grating has been superseded now by a technique of laser etching to produce a pattern of grooves in a concave glass surface. This concave holographic grating has the advantage of providing both the dispersing and the collecting mechanisms into a single component. The dispersed light can then be imaged or projected onto an array of photo-diodes. The concave grating requires less optical space and when combined with fiber optics, the instrument can be made extremely small and lightweight. This optical design is used in the Datacolor Spectraflash, Dataflash, and Microflash instruments. Detectors Just as gratings have improved in performance due to microprocessor technology, the detector assemblies have undergone similar revolutionary advancement. While the later version of the Hardy, and the Diano Match-Scan utilized the conventional analytical grade photomultiplier tubes (PMT), this highly accurate detector required the use of a moving slit to bring monochromatic light to the detector. Most modern instruments use fixed gratings and an array of photo-diode detectors to achieve the same purpose, but at a much lower cost of production and lower cost of maintenance in the long term. The manufacturing of micro-processors and integrated circuits has resulted in the development of high quality photo-diodes built on a single solid state electronic micro-chip. These silicon based diodes are ideal when placed in an array across the spectral distribution from a fixed diffraction grating or filter assembly. This optical assembly consisting of both grating and detector is referred to as the spectrometer shown in Figure 8. Because the optical components are fixed, these instruments are extremely stable exhibiting very little short-term or long-term drift in accuracy or precision. Another advantage is that these gratings and integrated detectors are highly reproducible. This has resulted in 6

7 LIGHT FROM SAMPLE DISPERSED LIGHT (REFERENCE BEAM) REFERENCE BEAM PHOTODETECTORS OUTPUT TO COMPUTER DISPERSED LIGHT (SAMPLE BEAM) DIFFRACTION GRATING LIGHT FROM REFERENCE SAMPLE BEAM PHOTODETECTORS OUTPUT Fig. 8 Dual Beam Spectrometer with twin detectors instruments which have excellent absolute agreement. This agreement between instruments is becoming much more important in global economies and many textile manufacturers and retailers are stipulating specific minimum color tolerances on goods to be shipped. Instrument Considerations Although many bench-top instruments are capable of measuring a diverse range of samples, there are some special considerations that the user should be aware of when choosing instruments. Some of these are covered in the next sections. Measurement of White Textiles Many textile companies produce white fabrics most of which is finished with fluorescent whitening agents (FWA) to achieve the desired bluish-white brightness. Many indices of whiteness have been developed which are suitable for measurement using any colorimeter or spectrophotometer. The AATCC recommended method is the C.I.E. Whiteness Index adopted as the AATCC Test Method 110. The method specifically states that the whiteness indices are relative and that the standard and batch are measured at about the same time on the same instrument. The reason for this is that in the case of fluorescent materials, of which FWA certainly qualifies, the emitted fluorescence is proportional to the overall intensity, or absolute number of photons, of the instrument illumination. Due to some variation in lamps, and the aging of such lamps, the whiteness values for FWA-brightened textiles tend to decrease throughout the life of the lamp or instrument. By always comparing the standard to the batch, this variation is usually not a problem provided the lamp has sufficient UV energy to excite the fluorescence. Many producers of white textiles prefer a more stable or absolute determination of whiteness. Instruments are available which are equipped with a filter calibrator for controlling the ratio of UV to visible output so that the illumination can approximate the distribution of standard illuminant D65. The method of calibration is that of Ganz and Griesser 5 and after illuminant calibration, the instrument is used to measure a more absolute whiteness, namely the Ganz Whiteness index. This method of illumination control is also useful for the measurement of visible fluorescent materials since the calibrated illumination leads to improved long-term repeatability. Instruments for Measuring Transmittance Most general purpose bench-top instruments have provision for the measurement of transmitted light as well as reflectance. The measurement of dyes in solution to verify the color quality and strength is the most common application, although the measurement of transparent films is also used. Most spectrophotometers for measuring liquids are designed such that a transmission cell or cuvet is inserted between the detector and the integrating sphere as shown in Fig 9. white tile lamp sphere total transmittance cell positions regular transmittance to spectrometer Fig. 9 Measurement positions for transmittance using Diffuse / 8 instrument The standardization for the measurement of tranmittance is generally performed by placing the white calibration tile at the sample reflectance port and setting the 100% transmittance with the solvent only in the cell. The zero (0%) transmittance is standardized by blocking the lens 7

8 or detector so that no light is allowed to enter the detector. The measurement of transmittance may be measured in two ways on sphere type instruments as either the total transmittance or regular transmittance. Total transmittance is measured by placing the cell flush against the sphere as shown in Fig. 9. In this way, the forward as well as side-scattered light is collected by the detector. In the other mode, the cell is positioned away from the sphere and closer to the detector. This measurement excludes all scattering except forward. In the measurement of transparent dyes in solution, the two methods yield identical results, however the total transmittance is most commonly used. Guide to Instrument Specifications The selection of an instrument for color measurement can be a rather confusing and time consuming task since there are many varieties in models with differing features and options. The following sections are given as a help to the colorist or lab manager in understanding some of the terms likely to appear on the technical brochures. More detailed descriptions of color terminology are given in ASTM E a Standard Terminology of Appearance. 6 Geometry - the angle of illumination / angle of detection in the optical system of the instrument Wavelength Range - total range (in nanometers nm) in which the instrument is capable of measuring, generally somewhere between nm, with most common. Bandwidth- in abridged or scanning spectrophotometers, the width of the measured band at 1/2 peak height used as a single point in the calculation and reporting of reflectance factors.. Bandwidths may range from 5nm - 20nm and is an important parameter in achieving good agreement between two instruments. Spectral Resolution - very similar to bandwidth, but indicates the actual spectral width being measured but not necessarily reported as a single point. An instrument may have diodes placed every 1nm however the data is integrated for every 10 diodes to give a bandwidth of 10nm but a spectral resolution of 1nm. This term is sometimes called sampling interval. Wavelength Accuracy - the average difference in nanometers between an instrument s working wavelength scale and the absolute scale as determined by the spectral emission lines from a discharge lamp. Photometric Accuracy - the accuracy in % Reflectance of the reflectance scale - usually 0-100% range. This is usally determined by measuring neutral tiles of known absolute reflectance. Performance Specifications Measurement speed - the time required to measure a sample including the time of actual data collection and the processing time to send the corrected data to the computer. Inter-instrument agreement - the average color difference expressed in either CIELAB de or CMC de between the instrument and a theoretical or real master instrument. This is usually determined in the factory by measuring a set of BCRA ceramic tiles on each instrument and calculating color differences from the master instrument. The BCRA tiles are suitable working standards although they are known to be thermochromic and manufacturers must work within controlled conditions. Repeatability - the color difference obtained when measuring a stable sample (usually a BCRA ceramic tile) repeatably on the same instrument, usually over a short period of time. Reproducibility - the color difference obtained when measuring a stable sample (usually a BCRA tile) over a longer period of time. This term usually includes variables such as time, operator, and conditions of the instrument. Some manufacturers report reproducibility as the color difference obtained on a single standard when measured on different instruments of the same model or type. Instruments for Special Purposes There are a variety of other insruments which have been developed for special purposes such as portability, continuous on-line measurement, goniophotometers, and extended wavelength instruments. A more complete description of instruments for the measurement of geometric and chromatic attributes of appearance is given by Hunter and Harold 7. 8

9 Portable Instruments The recent advances in integrated electronics and smaller optical components have lead to another revolution in color technology - the portable instrument. The variety in models and geometries are as diverse as in the benchtop models. Besides being completely portable, their attraction is that they can meet most quality control requirements without the use of an accompanying computer system. Their micro-processors are capable of calculating color differences, pass/fail, shade sorting, whiteness, grades of fastness, and many other indices of color and appearance. Their simplicity and lower cost relative to bench-tops have resulted in widespread use in quality inspection areas, retail, fabric and garment sourcing, and other areas which were essentially not using instruments and numerical methods previously. While these advantages have provided many users with the opportunity to now use color instrumentation, one must be aware of some limitations. Portable colorimeters, as with all colorimeters, are not capable of detecting metamerism. Many portables do not meet the same performance specifications as bench-top models in areas such as spectral resolution, bandwidth, and largeto-small viewing areas. Due to their size, many other features of bench-tops are not available, such as transmission measurement, and adjustable UV filters. Another key difference is sample measurement and presentation, a factor worth considering in textiles. Every bench-top instrument uses a sample holder such as a spring loaded plunger or air cylinder which applies a consistent amount of pressure to the back of the sample. This reduces the variability in measurement and thus lowers the number of reads required to achieve acceptable repeatability. Since most portables are handheld while measuring, the variability in pressure leads to higher variability in measurement. This is especially true when measuring fabrics or materials which have texture, pile, or are multi-layered allowing some pillowing of the fabric when measuring. Reversible Optics This term refers to an instrument which is capable of measuring in two modes - polychromatic illumination or monochromatic illumination. The Diano Match-Scan is the best example of this type and consists of two lamps and two detectors which allow both diffuse/0 and 0/diffuse geometries within the same instrument. The 0/diffuse mode is useful in measuring dye solutions or opaque samples which are light sensitive. The reversible optics also allow for the measurement of fluorescence by the two-mode method 8. Goniophotometers This instrument is designed for measuring reflectance at various and sometimes selectable viewing angles. They are used in areas where there is surface or internal scattering which changes the reflectances depending upon viewing angle. Examples in textiles are pigmented fibers and pile fabrics. On-Line Continuous Instruments Instruments for on-line color monitoring are used in carpeting and other continous wet processing, and are finding their way into the inspection area as well. These instruments are usually quite different from the benchtop models in the lab or dyehouse office. The most obvious difference is that most measure color without physical contact with the fabric since they are positioned above the web from 3 inches to 8 feet. On-line instruments must be very robust and capable of withstanding production environments as well as measuring while being traversed across the width of the frame. Concerning measurement geometry for on-line, it is not practical to use an integrating sphere except as a reference beam. Most instruments are bi-directional and actual illumination and viewing angles are determined by the position of the instrument when mounted above the plane of the fabric. For this reason, it is usually accepted that on-line measurements will not agree with measurements made off-line on samples taken to the lab, unless artificial correlation methods are applied. On-line measurement offers many advantages in color control such as real-time data allowing for immediate adjustment of pad roller pressure to correct for sidecenter-side variation. When used for monitoring, the system may be linked with a yardage meter and traversing frame to provide detailed color mapping of the roll of fabric or carpeting. Since the temperature of the fabric may be variable as it exits from a dryer, a pyrometer may be required which will allow for correlation between production conditions and some standard environment. 9

10 Extended Wavelength Instruments Instruments have been developed for measuring reflectance and transmittance at wavelengths other than the visible ( nm). Reflectance in the near infrared region of nm is of interest to those providing textiles for military use such as uniforms, tents, and vehicle fabrics. Although not visible, this reflected light is detectable using infra-red sensitive photography and filters. Instruments which are capable of measuring nm are equipped with a special grating or interference filters. Likewise, those involved in measuring fluorescent whites may want to measure the near ultraviolet reflectance or transmittance below 400nm. For liquids, a precision UV/VIS analytical spectrophotometer is recommended. For reflectance measurement below 400nm, some manufacturers provide measurement down to about 350nm. If the instrument uses a tungsten filament lamp, a supplemental UV lamp, such as a deuterium gas lamp, may be installed to provide the required output in the UV region. Conclusion The words of Ruth Johnston-Feller 9 in 1979 still hold true today in that a whole generation of instruments is available with far greater speed, better precision, better short-term repeatability and more flexibility in application. These high standards of performance in instrumentation, the powerful new software programs, and increased flexibility continue to provide tremendous tools toward achieving total color control. As a result, those companies who have invested in color instrumentation in the past, as well as those who will invest in the future, are certain to reap the rewards of effective color measurement and control. References 1. Derby, Roland E. Jr, (1983), Color Technology in the Textile Industry, AATCC, Research Triangle Park, NC. 2. Hardy, Arthur C., (1935), Journal of Optical Society of America, Vol. 25, p Commission International E clairage, (1986), Colorimetry, Publication Newton, Sir Isaac, (1730), OPTICKS, Reprinted by Dover Publications, New York, N.Y., Fourth Edition 5. Griesser, Rolf, (1981), Rev. Prog. Coloration, Vol 11, p American Society for Testing and Materials (ASTM), Publication ASTM Standards on Color and Appearance Measurement, (1994) 7. Harold, Richard and R.S. Hunter (199?), The Measurement of Appearance, 2nd Edition John Wiley and Sons, New York, N.Y. 8. Simon, Fred T., (1972), Journal of Color Appearance, Volume 1 9. Johnston-Feller, Ruth, (1979), Color Technology in the Textile Industry, AATCC, Research Triangle Park, N.C. 10

11